Expanded HIV testing may find more acute infections as well as more late presenters

Expanding HIV testing into local hospitals and primary care may not only find more people who have lived with HIV for a long time without knowing it, but may also find more people who have only recently become infected, a pilot study from Vancouver in Canada has found. 

The study was presented in Vancouver itself, at last week’s 2014 Treatment as Prevention Workshop there. The workshop’s chair, Dr Julio Montaner, told delegates that the province of British Columbia, which has Vancouver as its largest city, has pioneered a policy of treating everyone with HIV on diagnosis, including its relatively high proportion of people who inject drugs. This has reduced HIV infections in the province from 18 per 100,000 in the general population (the highest in Canada) to 7 per 100,000, where some other Canadian provinces have seen increases.

Missed opportunities for testing

For treatment as prevention to work, the proportion of people with HIV who are diagnosed and know their status must increase. Dr Réka Gustafson, the Medical Director of Communicable Disease Control at Vancouver Coastal Health, told the workshop that until recently, although testing had been expanding, the proportion of people who were late presenters had not decreased. While the proportion with a CD4 count below 500 cells/mm³ had decreased from 73% in 2003 to 50% in 2011, and the proportion in acute infection had increased from less than 5% of the total to over 20%, the proportion diagnosed with a CD4 count below 200 cells/mm³ had remained at about 20% since 2005.

A separate study also presented at the conference (Cescon) showed that people living with HIV in British Columbia had considerably more medical appointments than HIV-negative people in the years before their diagnosis: undiagnosed people with HIV were more likely to visit a primary care physician – especially people in their 20s, who were twice as likely as their contemporaries to see a doctor and whose primary care attendance rates nearly matched those of HIV-negative people in their 50s. People with HIV were also twice as likely to be admitted to hospital than a random sample of the general population. One reason for this is that they were more likely to be people who injected drugs (PWIDs), but even non-PWIDs were 26% more likely to be admitted to hospital. All these appointments and admissions represent lost opportunities for an HIV test.

The testing pilot

A pilot project evaluating a programme of expanded HIV testing took place between October 2011 and June 2013. This involved instituting routine HIV testing for all admissions to the three acute care hospitals serving Vancouver and also routine testing of patients in primary care: in the latter case over 500 primary care physicians signed up to the pilot scheme.

The scheme resulted in an expansion of tests in hospitals from about 500 a month before October 2011 to 2500 a month in October 2012, and an expansion of tests in primary care from 650 a month during 2011 to 2000 a month in the first half of 2013. There were over 73,000 tests performed in the last six months of the pilot compared with 38,000 in any six-month period in 2008-10, a near-doubling.

The number of people diagnosed with HIV in hospitals doubled during the pilot, from 11 in 2010 to 30 in 2012 and 27 in 2013. There was not a similar increase in positive diagnoses in primary care – there were 28 in 2010 and 25 in 2012 – but HIV prevalence in this group is lower than in people admitted to hospital. While the 0.2% of people diagnosed in primary care, and in hospital outpatients, roughly matches the prevalence of HIV in the general population in British Columbia, the proportion diagnosed in hospital admissions was 0.5%.

Looking solely at the hospital pilot, 44% of all patients were offered a test and 36% actually had one. Only 6% of patients offered a test refused it.

People testing positive in the hospital pilot were twice as likely to be heterosexual compared to people in other testing centres (27 versus 14%) and three times more likely to have injected drugs (17 versus 6%). They were less likely to be gay men (48 versus 72%).

As expected, the hospital pilot picked up on considerably more late presenters: the proportion diagnosed with a CD4 count below 200 cells/mm³ was 35%, compared with 12% in other centres. However, and unexpectedly, it also detected more people in acute infection: these formed 25% of the total compared with 15%.

This may be because people with acute HIV may not suspect their symptoms have anything to do with HIV, especially if they have recently tested negative. Dr Gustafson gave two complementary case examples. In one, a man aged 51 was diagnosed with a CD4 count of 10 cells/mm³ and a viral load of 1.5 million copies/ml. He had had weight loss and eating problems and had seen a doctor eight times with symptoms since 2009, each representing a lost opportunity to diagnose. In the other, a man aged 31 arrived at the emergency department with abdominal pain. He had tested HIV negative only six weeks before, but was found to have acute HIV infection with a viral load over the limit of quantification of the test – more than 10 million copies/ml.

Conclusions

As a result of the pilot, the British Columbia health department has now moved to a policy whereby all people between 18 and 70 will be offered an HIV test every five years, with annual tests for those in higher-risk populations: even people over 70 will be offered one test. Anyone presenting with a worsening medical condition that needs laboratory tests will be tested for HIV too.

Dr Gustafson commented that physicians had to be “untrained” in the use of risk assessment to decide whether someone should be tested for HIV. Risk assessment stigmatises testing, she said, and blinds doctors to minorities, such as the 12.2% of people with HIV in Vancouver who are people who inject drugs; many people do not know they have been at risk or are not prepared to disclose it. And in British Columbia, where sexual health services are part of primary care, fewer than 25% of people diagnosed with another sexually transmitted infection were tested for HIV within three months of their diagnosis.  Moving towards routine tests would therefore detect a considerable number of previously unsuspected cases of HIV.